Self-standing hollow Ni-doped Mo₂C nanotube arrays induced by the Kirkendall effect for an efficient hydrogen evolution reaction in acidic and alkaline solutions

Abstract

The development of efficient nonprecious catalysts for the hydrogen evolution reaction (HER) in water electrolysis is highly desirable but challenging. Molybdenum carbides, as a promising candidate, show excellent catalytic activity in both acidic and alkaline solutions. In this work, we report a self-supporting hollow Ni-doped Mo₂C nanotube array with varying Ni concentrations. Specifically, a thin layer of Ni and C composites was deposited on the surface of MoO₃ by plasma-enhanced chemical vapor deposition (PECVD), and in the subsequent carbonization process, a hollow structure was formed due to the Kirkendall effect, which endows the obtained electrode with a high specific surface area and a superior superhydrophilic/superaerophobic surface. In addition, the incorporation of Ni into Mo₂C could weaken the H* adsorption and thus improve the HER catalytic activity, as confirmed by density functional theory calculations. As a result, the as-prepared Ni-doped hollow Mo₂C hexagonal prism arrays are endowed with high catalytic activity in both acidic and alkaline solutions, 93 mV in 1 M KOH and 122 mV in 0.5 M H₂SO₄ to drive 10 mA cm⁻². Our work may provide a new way to enhance the performance of nonprecious electrocatalysts for the HER.

---

Introduction

Hydrogen (H₂) as a promising secondary energy source has attracted tremendous attention due to its high energy density and eco-friendly features.^1,2 Currently, reforming traditional hydrocarbons like diesel, methane, and methanol, as well as the process of coal gasification, are the traditional and dominant methods to produce H₂. Unfortunately, it is inevitable that these processes generate substantial environmentally unfriendly by-products, such as carbon monoxide (CO) and carbon dioxide (CO₂).³ Recently, renewable energy-driven electrocatalytic water splitting has offered a potential way to produce green hydrogen for sustainable energy conversion.^4,5 In this regard, noble metal-based materials exhibit high catalytic activity to meet the requirements, but their high cost hinders large-scale applications.^6–8 Therefore, the realization of a hydrogen economy to develop nonprecious materials with high catalytic activity for the hydrogen evolution reaction (HER) is desirable.

In recent decades, molybdenum carbides have attracted tremendous attention as promising candidates due to their similar d-band electronic structure to Pt and high performance in both acidic and alkaline electrolytes.^9–13 However, the HER catalytic activity of Mo₂C is still inferior to that of Pt-based materials and it is unable to meet the requirements of highly efficient energy conversion. Theoretical studies have shown that the high density of empty d-orbitals in Mo atoms in pure Mo₂C results in a relatively negative hydrogen binding energy (ΔG_H*), which leads to the strong formation of Mo–H bonds on its surface and hinders the desorption process of active hydrogen atoms.^14–17 Modulating the electron structure through doping heteroatoms or constructing hetero-interfaces has been verified as an effective strategy to alter the electronic coordination around active sites and improve the catalytic activity.^18–21 Transition metals, such as Fe, Co and Ni, are particularly effective in optimizing the d-band electronic structure due to their weak M–H bond strengths, resulting in proper adsorption/desorption of H₂.¹⁶ This approach has been shown to be effective in previous studies. For instance, self-supported Co/Mo₂C carbon nanofibers are obtained via the electrospinning method following a carbonization process. The introduction of Co has been found to tune the electronic structure of active sites, leading to enhanced HER catalytic activity, 128 mV at 10 mA cm⁻² in 1 M KOH.²² We fabricated a Ni/Mo₂C HER catalyst with ultrafine particles embedded on carbon fibers via annealing metal salt precursors covering carbon fibers by the dip-coating method.²³ In addition to the design of the components, the control of the surface structure of the catalysts is of equal importance for the enhancement of HER performance. Currently, most of the reported Mo₂C catalysts are in powder form and embedded in carbon matrices, which leads to several drawbacks, including high series resistance, blocked active sites, and agglomeration. From another perspective, the nanoarray structure may facilitate the faster release of the as-produced gas bubbles from the catalyst surface, thereby accelerating the catalytic process occurring on the electrode.^24–26 Constructing an integrated electrode with a modulated nanostructure remains a challenge due to the inevitable high-temperature carbonization process.

In this work, we have successfully synthesized a series of Ni-doped Mo₂C nanotube arrays on carbon cloth as efficient self-supported HER catalysts, which was achieved through a carbonization process occurring from the outside to the inside. First, a MoO₃ hexagonal prism array on carbon cloth is obtained via the hydrothermal method. Then a layer of composites consisting of Ni and C is deposited onto the surface of MoO₃ prisms via plasma-enhanced chemical vapour deposition (PECVD), using nickelocene as the nickel source and CH₄ as the carbon source. The carbonization reaction initiates at the interface of the Ni&C composite and MoO₃ at high temperatures, causing Mo and Ni&C to move towards each other in opposite directions. Finally, a hollow nanotube array of Ni-doped Mo₂C is formed due to the Kirkendall effect. The samples obtained show a surface that is superhydrophilic and underwater superhydrophobic, which enhances mass transfer and bubble release. In addition, the concentration of Ni in Mo₂C can be controlled by adjusting the ratio of the Ni source during deposition. Compared to pure Mo₂C (M-000), the optimized hollow Ni-doped Mo₂C electrocatalyst with 8.86% Ni dopants (M-300) demonstrated the highest HER catalytic activity. It only required 93 mV in 1 M KOH and 122 mV in 0.5 M H₂SO₄ to drive 10 mA cm⁻² HER current density, respectively. The Density Functional Theory (DFT) calculations also show that the catalytic activity increases and then decreases with more Ni dopant introduced. This work presents a new method for fabricating various metal carbides with modulated nanostructures by utilizing the Kirkendall effect.

Experimental section

Preparation of self-supported MoO₃ hexagonal prism arrays

Self-supported MoO₃ hexagonal prism arrays on carbon cloth were fabricated by a seed-assisted hydrothermal method. Initially, 10 g of Na₂MoO₄·2H₂O was dissolved in 16 mL of deionized (DI) water, followed by the addition of 4 mL of HCl (37 wt%) to the solution. After stirring for 30 min, carbon cloth bundles (100 mg) were immersed in the solution for 10 min and then blow-dried using an air blower. Subsequently, the dried carbon cloth bundles were heated at 350 °C for 30 min to form MoO₃ nanoparticles. The prepared carbon cloth bundles were placed in a Teflon-lined stainless steel autoclave (100 mL) containing 2 g of (NH₄)₆Mo₇O₂₄, 5 mL of HNO₃ (65 wt%), and 75 mL of DI water. Following this, the autoclave was heated to 120 °C for 5 min and then allowed to cool to room temperature. The prepared MoO₃ hexagonal prism array was washed with deionized water and subsequently dried in a vacuum at 60 °C for 2 h.

Preparation of C/Ni/Mo and Ni-doped Mo₂C

The prepared MoO₃ hexagonal prism arrays were subjected to radio frequency (RF) plasma treatment at room temperature to deposit nickel (Ni) sources. The masses of nickelocene as a Ni source in the plasma were 0, 30, 80, 130 and 180 mg separately. The obtained samples were named M-t (t values of 0, 1, 2, 3, and 4 represent the original masses of nickelocene of 0, 30, 80, 130, and 180 mg, respectively). RF plasma discharge was performed at 500 W and 13.56 MHz for 2 min. Subsequently, carbon deposition was conducted via CH₄ plasma, with CH₄, H₂, and Ar flow rates of 5, 10, and 20 mL min⁻¹, respectively. RF plasma discharge was performed at 500 W, 500 °C, and 13.56 MHz for 8 min. Following the PECVD strategy, C/Ni/Mo containing Ni and C was obtained, where the samples were denoted as M-t0. C/Ni/Mo was then carburized at 800 °C under an Ar atmosphere at a ramp rate of 10 °C min⁻¹ and held at the final temperature for 3 h. The Ni-doped Mo₂C samples were denoted as M-t00.

Preparation of Pt/C working electrodes

5 mg of commercial Pt/C catalysts was first ultrasonically dispersed in a mixed solution containing 760 μL of isopropyl alcohol, 200 μL of water, and 40 μL of Nafion solution (5 wt%). This process was continued until a homogeneous electrocatalytic ink was successfully formed. Then, 20 μL of the ink was carefully applied to a carbon cloth electrode (5 mm in diameter, with an approximate area of 19.625 cm²) and allowed to dry in air. Consequently, the mass or density of the loaded electrocatalyst on the carbon cloth was estimated to be approximately 0.50 mg cm⁻².

Characterization

The samples were characterized for morphology and microstructure by scanning electron microscopy (SEM, JEOL JSM-6700F) and transmission electron microscopy (TEM, JEOL JEM-2100F) equipped with an Oxford energy-dispersive X-ray analysis system. X-ray diffraction (XRD) analysis was performed using a DX2700 diffractometer with a Cu Kα radiation source (λ = 0.15406 nm). For X-ray photoelectron spectra (XPS), an ESCALAB 250 electron spectrometer with an Al Kα radiation source (hν = 1486.6 eV) was used. In addition, Raman spectra were obtained using a micro-Raman spectrometer (Renishaw) equipped with a 532 nm laser at 0.2 mW. The water contact angle and the underwater bubble contact angle were determined using an OCA 15EC instrument from DataPhysics. Optical images of escaping bubbles on the electrode surface were obtained using a high-speed camera.

Electrochemical measurements

Electrochemical measurements for the HER were performed on a CHI760E electrochemical workstation using a typical three-electrode configuration at room temperature. The self-supporting samples were used directly as the working electrode, while the carbon rod and Ag/AgCl electrodes served as the counter and reference electrodes, respectively. Cyclic voltammetry (CV) tests were conducted until a stable electrocatalytic performance of the catalysts was achieved at a scan rate of 50 mV s⁻¹. Then, linear sweep voltammetry (LSV) tests were performed at 5 mV s⁻¹ in 1.0 M KOH and 0.5 M H₂SO₄ to evaluate the electrochemical performance. The electrochemically active surface area (ECSA) of the prepared catalyst was investigated by calculating the double-layer capacitance (C_dl) by recording the CV curves at different scan rates (10–150 mV s⁻¹) in 1.0 M KOH (range: −0.9 to −0.8 V vs. Ag/AgCl) and 0.5 M H₂SO₄ (range: −0.1 to 0 V vs. Ag/AgCl). In addition, electrochemical impedance spectroscopy (EIS) measurements were performed at an initial current density of 10 mA cm⁻², across a frequency range from 100 kHz to 0.01 Hz. Stability tests were conducted at 10 mA cm⁻² for 12 h. Notably, all the potentials in the measurements were calibrated against a reversible hydrogen electrode (RHE) using the Nernst equation to ensure accuracy and consistency.

E_{(vs. RHE)} = E_{(vs. Ag/AgCl)} − iR + 0.059 × pH + 0.197 (1)

where i refers to the current and R is the resistance.

DFT methods

The Vienna ab initio simulation package (VASP) was used to perform spin-polarized DFT calculations in this study. The projector-augmented wave (PAW) method was utilized to calculate the interaction between the ions and electrons, and the exchange–correlation energy was described using the Perdew–Burke–Ernzerhof functional. The convergence criteria for force and total energy were set at 0.02 eV Å⁻¹ and 10⁻⁵ eV, respectively, while calculation details were established using a 500 eV cutoff energy for the plane-wave basis sets. For the first Brillouin zone, a 3 × 3 × 1 Γ-centered Monkhorst–Pack k-point grid was adopted, and van der Waals (vdW) interactions were addressed using the DFT-D3 method in Grimme's scheme. Bader charge analysis was performed to analyse the charge distribution. The Gibbs free energy was calculated with the computational hydrogen electrode (CHE) model. The reaction free energy change (ΔG) for each elementary step is calculated as

ΔG = ΔE + ΔE_zpe − TΔS (2)

where ΔE is the energy difference, and ΔE_zpe and ΔS are the zero-point energy and entropy corrections between the adsorbed state and the gas phase, respectively. The d-band center was calculated using the formula:^27,28where x and ρ(x) are the electronic energy and electronic density of states, respectively. The integral domain is from the minimum energy to the maximal energy of the d electrons.

Results and discussion

The overall synthesis procedure for Ni-doped Mo₂C hollow nanotube arrays is illustrated schematically in Fig. 1a. Initially, a smooth hexagonal prism with a diameter of approximately 5 μm is fabricated using a seed-assisted hydrothermal method. Multiple prisms are then uniformly arranged on the surface of a carbon cloth to form the array morphology (Fig. S1a†). Subsequently, a layer of Ni&C composite prisms is uniformly deposited on the surface of MoO₃via PECVD, resulting in a notably rough surface, as shown in Fig. S1b.† In order to provide sufficient carbon for the subsequent carbonization reaction, carbon deposition is carried out with CH₄ as the carbon source, and numerous vertical nanosheets are acquired on the surface of the hexagonal prism as shown in Fig. S1c.† The characteristic bands of M-30 in Fig. S2† at 1592.14 cm⁻¹ and 1347.87 cm⁻¹ can be attributed to the G and D bands, respectively, which confirm the presence of carbon. After calcination at 800 °C for 3 h, the hexagonal prisms in M-30 evolved into a hollow nanotube array (Fig. 1b). Similarly, M-000 is also a hollow tubular array structure with rough surfaces (Fig. S3†). As illustrated in Fig. 1c, the TEM images of M-300 also demonstrate the tube morphology and rough surface. Moreover, there is no residual carbon on the surface of M-300 from the edges. The lattice fringes of 0.259 and 0.238 nm can be indexed to the (100) and (002) planes of Mo₂C (JCPDS No. 35-0787), respectively, confirming complete carbonization. Furthermore, the elemental mappings demonstrate the successful introduction of Ni into M-300, indicating that Ni atoms are doped with Mo₂C.

Fig. 1 Materials synthesis and characterization. (a) Schematic of the preparation of Ni-doped Mo₂C electrodes with plasma assistance. (b) SEM images of M-300. (c) TEM images and the corresponding EDS mapping of M-300. (d) XRD patterns of M-0, M-3, and M-300. (e) Mo 3d spectrum of M-300. (f) Ni 2p spectrum of M-300.

The phase and composition of all samples at different steps were investigated by XRD. As illustrated in Fig. 1d, the diffraction peaks of M-0 at 25.94°, 29.33°, and 35.46° can be attributed to the (210), (300), and (310) diffraction planes of MoO₃ (JCPDS, No. 21-0569), respectively. The peaks in M-3 are consistent with those in M-0, indicating that the plasma deposition process does not change the phase of the molybdenum oxide prisms. Following the carbonization at 800 °C for 3 h, the XRD pattern of M-300 exhibited sharp peaks at 34.34°, 38.14°, and 39.31°. These peaks are indexed to the (100), (002), and (101) planes of Mo₂C (JCPDS, No. 35-0787), while the characteristic peaks of MoO₃ are absent, indicating a complete phase transformation from MoO₃ to Mo₂C. Furthermore, the XRD patterns of M-100, M-200, and M-400 also exhibit typical characteristic peaks of Mo₂C, suggesting that the variation in Ni content does not impact the primary phase of Mo₂C (Fig. S4†). To further investigate the composition and valence state of the samples, XPS analysis was performed. The peak at 857.1 eV in the full scan spectrum of M-3 and M-300 indicates that Ni element was successfully introduced via the PECVD method and carbonization process (Fig. S5†). As shown in Fig. 1e, the Mo 3d spectrum of M-300 reveals the existence of four distinct Mo species, namely Mo²⁺, Mo³⁺, Mo⁴⁺, and Mo⁶⁺. The Mo²⁺ and Mo³⁺ species are associated with the Mo–C in the Mo₂C coating, while the Mo⁴⁺ and Mo⁶⁺ species are attributable to the presence of MoO_x, which is formed when Mo₂C is exposed to air.^29–31 As illustrated in Fig. 1f, the fitted curve of the Ni 2p XPS spectrum reveals that the peaks at 855.6 and 873.0 eV are attributed to Ni²⁺, with satellite peaks situated at 861.5 and 879.1 eV, corresponding to Ni–C.^19,32 Furthermore, as the mass of the Ni source utilized in the plasma increases, the peak signal of the Ni element in the fine spectrum becomes progressively stronger (Fig. S6†). The ratio between Ni and Mo in M-100, M-200, M-300, and M-400 was calculated according to the areas of the fine spectrum. The results indicate that an increase in the mass of the Ni source results in greater Ni doping in Mo₂C (Table S1†). As illustrated in Fig. S7,† the spectrum of M-300 exhibits several prominent Raman peaks at 149.36, 280.35, 351.09, 774.76, 852.00, and 994.56 cm⁻¹, which can be attributed to Mo₂C.^33–35 Notably, no characteristic peaks for carbon are detected, indicating that there is no excess carbon, which is consistent with the aforementioned TEM results.

The structural evolution of Ni-Mo₂C hollow nanotubes was investigated in detail by studying the annealing process at 800 °C for different periods of time (1 h, 2 h, and 3 h). When exposed at 800 °C, the reaction between MoO₃ and carbon is initiated, as shown in Fig. 2a. After calcination at 800 °C for one h, the carbon nanosheets covering the MoO₃ surface gradually disappeared while the particle structure was gradually formed. This change highlights the initial stages of carbonization occurring at the MoO₃/carbon interface. As the calcination time increases, the surface roughness increases, indicating the occurrence of internal deformations (Fig. 2b). These observations suggest a mutual diffusion process between carbon and MoO₃. The XRD analysis reveals distinct phases of the product based on the duration of calcination (Fig. S8†). Initially, after 1 h of calcination, the product phase is identified as MoO₂. This progression highlights the reduction of MoO₃ to MoO₂ through carbon in the initial stages. With an extended calcination period of 2 h, the product phase undergoes a transition, becoming a combination of MoO₂ and Mo₂C. Following a further 3 h of calcination, MoO₂ undergoes a gradual transformation into the final product, Mo₂C. Combining this with previous reports,³⁶ the following chemical equations are provided to describe the conversion process:

MoO₃ + 0.5C = MoO₂ + 0.5CO₂ (4)

MoO₂ + 2.5C = 0.5Mo₂C + 2CO (5)

Fig. 2 Formation mechanism of hollow tubes. (a–c) SEM images of M-30 calcined at 800 °C for different time periods. (d) Schematic illustration of structural evolution from 1 h to 3 h. (e) Formation mechanism of Mo₂C hollow nanotubes.

As illustrated in Fig. 2c, the diffusion pattern intensifies as the calcination period progresses, resulting in an enhanced reactivity between carbon and molybdenum and the gradual formation of a hollow tubular structure. In light of the specific evolution of Mo₂C phases at varying annealing time periods, it is postulated that the time-dominant Kirkendall effect may provide a reasonable explanation for the formation of hollow tubes, as schematically illustrated in Fig. 2d.^37,38 At 800 °C, Mo⁶⁺ is reduced to Mo²⁺ through carbothermal reduction. The formation of Mo²⁺ is visually demonstrated in Fig. 2e, where it diffuses outward rapidly, whereas atomic C diffuses inward slowly, resulting in the formation of the Mo₂C phase. The disparity in their diffusion rates leads to the formation of a cavity in the inner region, which subsequently evolves into a larger central cavity, ultimately resulting in the hollow nanotube structure.^39,40 The structure provides Ni-doped Mo₂C with numerous active sites, three-phase interfacial interactions, and efficient mass transport, which are beneficial for the HER.

The electrochemical performances of self-standing Mo₂C-based electrodes were evaluated in both 1 M KOH and 0.5 M H₂SO₄ electrolytes. As illustrated in Fig. 3a and Fig. S9a,† the M-300 electrode exhibits a geometric current density (j) of 10 mA cm⁻² in 1 M KOH solution with only a small overpotential of 93 mV, which is lower than that of the M-100 (147 mV), and M-400 (124 mV) electrodes. In particular, the overpotential of M-300 is comparable to that of M-200. At current densities of 10 mA cm⁻² and 100 mA cm⁻², the overpotentials for M-000 are 154 mV and 234 mV, respectively, which are higher than that of M-300. These values indicate that the catalytic activity of Mo₂C is significantly enhanced after Ni doping. The Tafel plots show that the M-300 electrode has the smallest Tafel slope value of 61.93 mV dec⁻¹ among M-000 (75.32 mV dec⁻¹), M-100 (75.26 mV dec⁻¹), M-200 (64.07 mV dec⁻¹), and M-400 (73.93 mV dec⁻¹) (Fig. 3b and Fig. S9b†), indicating that M-300 exhibits superior kinetics for the HER. These results suggest that the HER activity of the as-prepared M-300 is superior to those of most previously reported Mo₂C-based HER electrocatalysts (Table S2†). Furthermore, the HER performances of the obtained Mo₂C-based electrocatalysts were also evaluated in acidic solution. At a current density of 10 mA cm⁻² in 0.5 M H₂SO₄, the overpotential values for M-000, M-100, M-200, M-300, and M-400 are 220, 185, 128, 122 and 183 mV, respectively (Fig. 3d and Fig. S9c†). The Tafel plots demonstrate that the M-300 electrode exhibits the lowest Tafel slope value of 68.55 mV dec⁻¹ among M-000 (88.04 mV dec⁻¹), M-100 (85.77 mV dec⁻¹), M-200 (72.05 mV dec⁻¹), and M-400 (85.62 mV dec⁻¹), which suggests that M-300 has the most favourable HER kinetics in acidic solution (Fig. 3f and Fig. S9d†). Moreover, the HER performance of the M-300 electrode exceeds that of other Mo₂C-based materials obtained by other reported HER catalysts in 0.5 M H₂SO₄ (Table S2†). The C_dl can be employed to further analyse the HER catalytic performance of Mo₂C-based electrocatalysts. In 1 M KOH, the C_dl value of M-300 (18.73 mF cm⁻²) is higher than those of M-000 (12.77 mF cm⁻²), M-100 (13.07 mF cm⁻²), M-200 (15.75 mF cm⁻²), and M-400 (15.45 mF cm⁻²). Moreover, the C_dl of M-300 in 0.5 M H₂SO₄ is also higher than those of the other electrocatalysts (Fig. S9e and S9f†). These results suggest that M-300 exhibits a larger electrochemical surface area compared to M-000, M-100, M-200 and M-400.

Fig. 3 HER performance. (a) The LSV curves and (b) the Tafel plots in 1 M KOH. (d) The LSV curves and (e) the Tafel plots in 0.5 M H₂SO₄. The LSV curves before and after stability tests of M-300 in 1 M KOH (c) and 0.5 M H₂SO₄ (e) (inset: stability curves of M-300).

High conductivity inevitably provides a faster charge transfer pathway for catalysed reactions. Consequently, EIS was employed to further characterize the interfacial charge transfer kinetics of the catalyst. Fig. S10a and S10b† demonstrate the resistive and capacitive characteristics of the as-prepared samples. The corresponding equivalent circuit diagram, which includes R_s (electrolyte resistance), R_ct (charge transfer resistance), and CPE (constant phase element), is shown in the inset. The radius of the semicircle at high frequency in the Nyquist plots corresponds to R_ct. A smaller R_ct indicates faster charge transfer kinetics, which results in a more efficient Volmer reaction. The M-300 electrode exhibits a lower charge transfer resistance than the M-000 electrode in 1 M KOH. Similarly, the M-300 electrode also exhibits smaller charge transfer resistance than M-000 in 0.5 M H₂SO₄. The Nyquist plots confirm the smaller charge transfer resistance of the M-300 electrode. The lower R_ct shows that the Ni doping in M-300 with a hollow nanotube array structure enhances the electron transfer rate and improves the conductivity of Mo₂C. Moreover, it is important to ensure the long-term stability of the catalyst.

The stability of M-000 and M-300 self-supporting electrodes for the HER in 1 M KOH and 0.5 M H₂SO₄ electrolytes was evaluated by conducting sequential electrolysis at 10 mA cm⁻² for 12 h. Following the test, the potential for M-300 increased by 34 mV after 12 h, which was lower than that for M-000 in 1 M KOH (Fig. 3c and Fig. S11a†). As depicted in Fig. 3c, the polarization curves of M-300 exhibited minimal changes after the stability test. Conversely, the potential of M-000 exhibited a more pronounced increase (Fig. S11b†). Similarly, when subjected to 12 h of stability testing in 0.5 M H₂SO₄, the potential for M-300 increased by 10 mV, which was also lower than that for M-000 (Fig. 4c and Fig. S12a†). The polarization curves of M-300 showed smaller changes compared to M-000. Specifically, the overpotential of M-300 at 10 mA cm⁻² increased by 16 mV, while that of M-000 increased by 29 mV (Fig. S12b†). The composition of M-300 was compared before and after the stability measurement in H₂SO₄ to investigate its remarkable stability. The distinctive peaks of Mo₂C remained unaltered following the stability tests, as evidenced by the XRD patterns (Fig. S13†). Furthermore, SEM images reveal that the structure of the hexagonal prism has undergone slight changes, but the hollow nanotube structure of M-300 remains unchanged (Fig. S14†). These results demonstrate good stability of M-300 for the HER.

Fig. 4 Contact angle images and bubble release behaviour on the electrode surface. (a) Water contact angles at different times in air for M-000 and M-300. (b) Bubble contact angles at different times in water for M-000 and M-300. (c) The bubble evolution process on the surface of M-000 and M-300 electrodes at 10 mA cm⁻² in 1 M KOH.

The surface properties are of great importance in determining HER performance. In particular, the gas–liquid–solid triphase is critical to the hydrogen generation process. The contact angles in 1 M KOH alkaline solution were measured to determine the surface properties of M-000 and M-300. As illustrated in Fig. 4a, no droplet shadow was captured using the high-speed camera on the surface of M-300 at 68 ms, indicating that the water droplets spread rapidly on the surface of M-300 at that point. In contrast, the water droplets on the M-000 surface were not completely dispersed at the same time. This difference in wetting behaviour indicates the superior superhydrophilicity of the M-300 surface. The superhydrophilic electrode promotes close contact with the electrolyte, enhancing both surface site availability and reaction speed. In addition to the droplet wettability measurements, the contact angle of underwater gas bubbles on the prepared electrode was also measured. As shown in Fig. 4b, the gas bubble contact angles were found to be 145.8 and 154.7° for M-000 and M-300 at 120 s respectively, suggesting that M-300 is more superaerophobic than M-000. This characteristic can be attributed to the discontinuous state of the three-phase contact line of the bubbles within the hierarchical surface of the M-300 electrode, which contributes to the extremely low contact region between the bubbles and the electrode surface.^41,42 The superaerophobic structure of M-300 is advantageous for rapidly releasing H₂ bubbles and exposing the active sites. A high-speed camera was also used to study H₂ bubble release behaviour from the electrode surface during the HER at current densities of 10 mA cm⁻² for M-000 and M-300. Fig. 4c shows that small bubbles (represented by white circles) were already present on the surface of M-000 at 80 s. Over time, these bubbles did not dissipate but instead increased in size. Interestingly, the surface of the M-300 electrode did not exhibit any bubble accumulation at 80 s and 120 s, which is similar to the surface of the electrode at time 0 s. This confirms that bubbles detach more easily from the surface of M-300 than from that of M-000. Superhydrophilicity and superaerophobicity could facilitate the access of the electrolyte to active sites during the HER for M-300. This may be one of the reasons for the excellent catalytic properties of M-300.

DFT calculations were conducted to explore the impact of Ni doping on the HER performance of Mo₂C (Fig. 5 and Fig. S15–S18†). 1-Ni, 2-Ni, 3-Ni, 4-Ni, and 5-Ni represent the number of Ni substitutions for Mo as 1, 2, 3, 4, and 5. An ideal HER catalyst should have a near-zero free energy of adsorbed H* (ΔG_H* ≈ 0). The ΔG_H* of pure Mo₂C (0-Ni) is too negative (−0.62 eV), indicating that pure Mo₂C is not suitable for the HER. As the Ni-doping concentration increases, the values of ΔG_H* on Ni-doped Mo₂C gradually tend towards zero (Fig. 5g). The ΔG_H* values of 1-Ni/2-Ni/3-Ni-doped Mo₂C are −0.58, −0.53, and −0.23 eV, respectively, which means that Ni doping could weaken the H* adsorption and thus improve the HER performance of Mo₂C. Notably, Ni-doping concentrations that are too high do not further improve the HER performance of Mo₂C, which can be seen from the ΔG_H* values of 4-Ni/5-Ni-doped Mo₂C (−0.25 and −0.52 eV, respectively). The trend of the DFT results is well consistent with our electrochemical tests, and the Bader charge analysis provides a deeper understanding of this trend. The charges (Δq) donated from 0-Ni/1-Ni/2-Ni/3-Ni/4-Ni/5-Ni-doped Mo₂C to H* are positively correlated with ΔG_H* (Fig. 5g and h), which shows that the greater the Δq, the stronger the H* adsorption. The 0-Ni-doped Mo₂C donates 0.5 e⁻ to H* causing strong H* absorption, which would not be beneficial for the subsequent HER steps. As for 3Ni-doped Mo₂C, the charge donated to H* is reduced to 0.28 e⁻, resulting in relatively moderate H* absorption. The change in H* adsorption on these two surfaces could also be reflected by the d-band center (ε_d), where the ε_d of Mo atoms in 0-Ni-doped Mo₂C is −0.81 eV, whereas the ε_d of Ni atoms in 3-Ni-doped Mo₂C is −1.77 eV, and the deviation of the Ni-ε_d from the Fermi level means weaker H* absorption. Hence, Ni-doping contents could adjust the charge transfer between the electrodes and H* species, thus improving the HER performance.

Fig. 5 Theoretical calculations. Ni-doped Mo₂C model diagrams of (a) 0-Ni, (b) 1-Ni, (c) 2-Ni, (d) 3-Ni, (e) 4-Ni, and (f) 5-Ni. (h) The ΔG and (g) Δq values for different Ni contents.

Conclusions

In summary, self-supporting hollow nanotube Ni-doped Mo₂C array electrodes were successfully produced by the Kirkendall effect for the HER in acidic and alkaline solutions. The M-300 electrode, featuring a hollow array structure with a large specific surface area, a superhydrophilic/superaerophobic surface and excellent charge transfer capability, exhibited a low overpotential of 93 mV in 1 M KOH, as well as 122 mV in 0.5 M H₂SO₄ at 10 mA cm⁻². Significantly, the electrode showed remarkable durability compared to the Mo₂C electrode without Ni doping. This study provides valuable insights for the advancement of high-performance Ni-doped carbide HER catalysts.

Author contributions

Chen Li: conceptualization, methodology, investigation, data curation, writing – original draft, and writing – review & editing. Beirong Ye: investigation and data curation. Tengfei Zhang: data curation. Renhong Chen: data curation. Yongqi Li: data curation. Xin Liu: funding acquisition. Tongwei Wu: funding acquisition, density functional theory calculations and data analyses. Hongxian Liu: funding acquisition, density functional theory calculations and data analyses. Xinhui Xia: funding acquisition. Yongqi Zhang: conceptualization, methodology, writing – review & editing, supervision, project administration, and funding acquisition.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

H. Liu thanks the Youth Science Foundation of Guizhou Province Education Ministry (QJJ[2022]318), the Doctor Foundation, and the Academic Training Project of Zunyi Normal University (ZSBS[2022]08 and ZSXM[2023]1-04) for their support. T. W. was supported by the National Natural Science Foundation of China (No. 52202214) and the Natural Science Foundation of Sichuan Province (No. 2023NSFSC0954). The numerical calculations in this paper were performed at the Hefei Advanced Computing Center. This work is supported by the National Natural Science Foundation of China (Grant No. 52073252, 52002052), the Science and Technology Department of Zhejiang Province (Grant No. 2023C01231), the Key Research and Development Project of the Science and Technology Department of Sichuan Province (2022YFSY0004), the Key Laboratory of Engineering Dielectrics and Its Application (Harbin University of Science and Technology), Ministry of Education (Grant No. KFM 202202) the Natural Science Foundation for Distinguished Young Scholars of Zhejiang Province (Grant No. LR20E020001), the Open Project Program of the State Key Laboratory of Photocatalysis on Energy and Environment (Grant No. SKLPEE-KF202206), Fuzhou University, and the Open Project Program of the State Key Laboratory of New Textile Materials and Advanced Processing Technologies (Grant No. FZ2021009). The authors appreciate Tingchuan Zhou from Analysis and Testing Center, University of Electronic Science and Technology of China, for technical support.

References

1. W. Ma, Z. Qiu, J. Li, L. Hu, Q. Li, X. Lv and J. Dang, Interfacial electronic coupling of V-doped Co2P with high-entropy MXene reduces kinetic energy barrier for efficient overall water splitting, J. Energy Chem., 2023, 85, 301–309.
2. K. Guo, J. Jia, H. Wang and C. Xu, Unveiling the surface hydroxylation and selenite modification of in situ generated nickel for promoting the hydrogen evolution reaction, Inorg. Chem. Front., 2024, 11, 3357–3366.
3. J. A. Turner, Sustainable hydrogen production, Science, 2004, 305, 972–974.
4. X. Jian, W. Zhang, Y. Yang, Z. Li, H. Pan, Q. Gao and H.-J. Lin, Amorphous Cu–W alloys as stable and efficient electrocatalysts for hydrogen evolution, ACS Catal., 2024, 14, 2816–2827.
5. J. Zhan, Y. Cao, J. Lai, J. Li, H. Ma, S. Li, P. Liu, X.-Q. Chen and Y. Sun, Design of high-efficiency hydrogen evolution catalysts in a chiral crystal, ACS Catal., 2024, 14, 1030–1036.
6. J. Wang, T. Yang, X. Li, H. Zhang, Y. Zhang, Y. He and H. Xue, Hydrogen evolution reaction activity enhancement from active site turnover mechanism, J. Energy Chem., 2024, 92, 629–638.
7. T. T. Yang, A. Wang, S. D. House, J. Yang, J.-K. Lee and W. A. Saidi, Computationally guided design to accelerate discovery of doped β-Mo2C catalysts toward hydrogen evolution reaction, ACS Catal., 2022, 12, 11791–11800.
8. L. Mei, Y. Zhang, Z. Ye, T. Han, H. Hu, R. Yang, T. Ying, W. Zheng, R. Yan, Y. Zhang, Z. Wang and Z. Zeng, Fabrication of amorphous subnanometric palladium nanostructures on metallic transition metal dichalcogenides for efficient hydrogen evolution reaction, Inorg. Chem. Front., 2024 10.1039/d4qi00622d.
9. H. Cheng, L. X. Ding, G. F. Chen, L. Zhang, J. Xue and H. Wang, Molybdenum carbide nanodots enable efficient electrocatalytic nitrogen fixation under ambient conditions, Adv. Mater., 2018, 30, 1803694.
10. J. Jia, T. Xiong, L. Zhao, F. Wang, H. Liu, R. Hu, J. Zhou, W. Zhou and S. Chen, Ultrathin N-doped Mo2C nanosheets with exposed active sites as efficient electrocatalyst for hydrogen evolution reactions, ACS Nano, 2017, 11, 12509–12518.
11. X. Lu, M. Li, H. Wang and C. Wang, Advanced electrospun nanomaterials for highly efficient electrocatalysis, Inorg. Chem. Front., 2019, 6, 3012–3040.
12. Y. Zhou, J. Xu, C. Lian, L. Ge, L. Zhang, L. Li, Y. Li, M. Wang, H. Liu and Y. Li, Carbon impurity-free, novel Mn,N co-doped porous Mo2C nanorods for an efficient and stable hydrogen evolution reaction, Inorg. Chem. Front., 2019, 6, 2464–2471.
13. J. Wang, S. Li, J. Hu, S. Niu, Y. Li and P. Xu, Acid-directed morphology control of molybdenum carbide embedded in a nitrogen doped carbon matrix for enhanced electrocatalytic hydrogen evolution, Inorg. Chem. Front., 2020, 7, 3620–3626.
14. Y. Xu, J. Yang, T. Liao, R. Ge, Y. Liu, J. Zhang, Y. Li, M. Zhu, S. Li and W. Li, Bifunctional water splitting enhancement by manipulating Mo–H bonding energy of transition metal–Mo2C heterostructure catalysts, Chem. Eng. J., 2022, 431, 134126.
15. G. Liu, K. Wang, L. Wang, B. Wang, Z. Lin, X. Chen, Y. Hua, W. Zhu, H. Li and J. Xia, A Janus cobalt nanoparticles and molybdenum carbide decorated N-doped carbon for high-performance overall water splitting, J. Colloid Interface Sci., 2021, 583, 614–625.
16. W. Liu, X. Wang, J. Qu, X. Liu, Z. Zhang, Y. Guo, H. Yin and D. Wang, Tuning Ni dopant concentration to enable co-deposited superhydrophilic self-standing Mo2C electrode for high-efficient hydrogen evolution reaction, Appl. Catal., B, 2022, 307, 121201.
17. C. Yang, R. Zhao, H. Xiang, J. Wu, W. Zhong, W. Li, Q. Zhang, N. Yang and X. Li, Ni-Activated transition metal carbides for efficient hydrogen evolution in acidic and alkaline solutions, Adv. Energy Mater., 2020, 10, 2002260.
18. J. Wang, T. Liao, Z. Wei, J. Sun, J. Guo and Z. Sun, Heteroatom-doping of non-noble metal-based catalysts for electrocatalytic hydrogen evolution: An electronic structure tuning strategy, Small Methods, 2021, 5, e2000988.
19. F. Yu, Y. Gao, Z. Lang, Y. Ma, L. Yin, J. Du, H. Tan, Y. Wang and Y. Li, Electrocatalytic performance of ultrasmall Mo2C affected by different transition metal dopants in hydrogen evolution reaction, Nanoscale, 2018, 10, 6080–6087.
20. H. Lin, N. Liu, Z. Shi, Y. Guo, Y. Tang and Q. Gao, Cobalt-doping in molybdenum-carbide nanowires toward efficient electrocatalytic hydrogen evolution, Adv. Funct. Mater., 2016, 26, 5590–5598.
21. G. Liu, H. Bai, Y. Ji, L. Wang, Y. Wen, H. Lin, L. Zheng, Y. Li, B. Zhang and H. Peng, A highly efficient alkaline HER Co–Mo bimetallic carbide catalyst with an optimized Mo d-orbital electronic state, J. Mater. Chem. A, 2019, 7, 12434–12439.
22. J. Wang, R. Zhu, J. Cheng, Y. Song, M. Mao, F. Chen and Y. Cheng, Co, Mo2C encapsulated in N-doped carbon nanofiber as self-supported electrocatalyst for hydrogen evolution reaction, Chem. Eng. J., 2020, 397, 125481.
23. Y. Zhang, X. Xia, X. Cao, B. Zhang, N. H. Tiep, H. He, S. Chen, Y. Huang and H. J. Fan, Ultrafine metal nanoparticles/N-doped porous carbon hybrids coated on carbon fibers as flexible and binder-free water splitting catalysts, Adv. Energy Mater., 2017, 7, 1700220.
24. X. Yu, Z.-Y. Yu, X.-L. Zhang, Y.-R. Zheng, Y. Duan, Q. Gao, R. Wu, B. Sun, M.-R. Gao, G. Wang and S.-H. Yu, “Superaerophobic” nickel phosphide nanoarray catalyst for efficient hydrogen evolution at ultrahigh current densities, J. Am. Chem. Soc., 2019, 141, 7537–7543.
25. H. Sun, B. Yao, Y. Han, L. Yang, Y. Zhao, S. Wang, C. Zhong, J. Chen, C. P. Li and M. Du, Multi-interface engineering of self-supported nickel/yttrium oxide electrode enables kinetically accelerated and ultra-stable alkaline hydrogen evolution at industrial-level current density, Adv. Energy Mater., 2024, 14, 2303563.
26. H. Sun, L. Li, M. Humayun, H. Zhang, Y. Bo, X. Ao, X. Xu, K. Chen, K. Ostrikov, K. Huo, W. Zhang, C. Wang and Y. Xiong, Achieving highly efficient pH-universal hydrogen evolution by superhydrophilic amorphous/crystalline Rh(OH)3/NiTe coaxial nanorod array electrode, Appl. Catal., B, 2022, 305, 121088.
27. A. Nilsson, L. G. M. Pettersson, B. Hammer, T. Bligaard, C. H. Christensen and J. K. Nørskov, The electronic structure effect in heterogeneous catalysis, Catal. Lett., 2005, 100, 111–114.
28. J. K. N. B. Hammer and J. K. Nørskov, Theoretical surface science and catalysis—Calculations and concepts, Adv. Catal., 2000, 71–129.
29. J.-S. Li, Y. Wang, C.-H. Liu, S.-L. Li, Y.-G. Wang, L.-Z. Dong, Z.-H. Dai, Y.-F. Li and Y.-Q. Lan, Coupled molybdenum carbide and reduced graphene oxide electrocatalysts for efficient hydrogen evolution, Nat. Commun., 2016, 7, 11204.
30. L. He, W. Zhang, Q. Mo, W. Huang, L. Yang and Q. Gao, Molybdenum carbide-oxide heterostructures: In situ surface reconfiguration toward efficient electrocatalytic hydrogen evolution, Angew. Chem., Int. Ed., 2020, 59, 3544–3548.
31. J. Staszak-Jirkovský, C. D. Malliakas, P. P. Lopes, N. Danilovic, S. S. Kota, K.-C. Chang, B. Genorio, D. Strmcnik, V. R. Stamenkovic, M. G. Kanatzidis and N. M. Markovic, Design of active and stable Co–Mo–Sx chalcogels as pH-universal catalysts for the hydrogen evolution reaction, Nat. Mater., 2015, 15, 197–203.
32. Z. Hu, L. Zhang, J. Huang, Z. Feng, Q. Xiong, Z. Ye, Z. Chen, X. Li and Z. Yu, Self-supported nickel-doped molybdenum carbide nanoflower clusters on carbon fiber paper for an efficient hydrogen evolution reaction, Nanoscale, 2021, 13, 8264–8274.
33. J. Wan, J. Wu, X. Gao, T. Li, Z. Hu, H. Yu and L. Huang, Structure confined porous Mo2C for efficient hydrogen evolution, Adv. Funct. Mater., 2017, 27, 1703933.
34. T. Li, W. Luo, H. Kitadai, X. Wang and X. Ling, Probing the domain architecture in 2D α-Mo2C via polarized raman spectroscopy, Adv. Mater., 2019, 31, 1807160.
35. S. Hussain, S. A. Zaidi, D. Vikraman, H.-S. Kim and J. Jung, Facile preparation of molybdenum carbide (Mo2C) nanoparticles and its effective utilization in electrochemical sensing of folic acid via imprinting, Biosens. Bioelectron., 2019, 140, 111330.
36. G. D. Sun, G. H. Zhang, B. J. Yan and K. C. Chou, Study on the reduction of commercial MoO3 with carbon black to prepare MoO2 and Mo2C nanoparticles, Int. J. Appl. Ceram. Technol., 2020, 17, 917–931.
37. X. Tang, J. Zhao, L. Zhou, S. Kuang, Z. Pi, H. Lin and Q. Li, Self-templating approach to Co9S8 hexagonal sheets via Kirkendall effect with efficient photocatalytic performance, Inorg. Chem. Commun., 2024, 161, 112055.
38. Y. Boyjoo, M. Wang, V. K. Pareek, J. Liu and M. Jaroniec, Synthesis and applications of porous non-silica metal oxide submicrospheres, Chem. Soc. Rev., 2016, 45, 6013–6047.
39. T. Gong, Y. Liu, K. Cui, J. Xu, L. Hou, H. Xu, R. Liu, J. Deng and C. Yuan, Binary molten salt in situ synthesis of sandwich-structure hybrids of hollow β-Mo2C nanotubes and N-doped carbon nanosheets for hydrogen evolution reaction, Carbon Energy, 2023, 5, e349.
40. Y. Yin, R. M. Rioux, C. K. Erdonmez, S. Hughes, G. A. Somorjai and A. P. Alivisators, Formation of hollow nanocrystals through the nanoscale kirkendall effect, Science, 2004, 304, 711–714.
41. X. Shan, J. Liu, H. Mu, Y. Xiao, B. Mei, W. Liu, G. Lin, Z. Jiang, L. Wen and L. Jiang, An engineered superhydrophilic/superaerophobic electrocatalyst composed of the supported CoMoSx chalcogel for overall water splitting, Angew. Chem., Int. Ed., 2019, 59, 1659–1665.
42. Y. Li, H. Zhang, T. Xu, Z. Lu, X. Wu, P. Wan, X. Sun and L. Jiang, Under-water superaerophobic pine-shaped Pt nanoarray electrode for ultrahigh-performance hydrogen evolution, Adv. Funct. Mater., 2015, 25, 1737–1744.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4qi01427h

---